The demand for ever shorter charging times for electric vehicles is currently one of the major challenges the automotive industry faces. The maximum achievable charging performance of the battery is limited by the physical properties of the battery cells. Accurate knowledge of the internal cell states offers the opportunity to operate Li-ion cells as close as possible to their physical limits without increasing the risk of accelerated ageing compared to the state of the art. Particularly in the case of fast charging, optimal utilization of the physical limits can so be expected to shorten the charging time. Since internal operating conditions such as the anode potential can hardly be measured during operation, modeling them using real-time capable electrochemical models is a promising approach.
In this work a Single Particle Model with Electrolyte (SPMe) is used. This is coupled with a thermal model, implemented as a lumped-parameter model. The coupling of the model with real sensor data from a battery cell is achieved by integrating the electrochemical model into an observer structure. The estimation accuracy of the observer can be improved by implementing a filter algorithm. In this work, an extended Kalman filter is used to update the state of the model in the case of a deviation between the model and the real cell. Therefore, the degree of lithiation of the cathode in the model is corrected based on the observed deviation in cell voltage, thereby updating the State of Charge (SOC).
The performance of the model is validated using real measurement data. Various SOC offsets between the model and the real cell are simulated. Based on the settling time and the accuracy of the simulated cell voltage and anode potential, the performance of the model is evaluated. To demonstrate real-time capability and also the potential of the methodology for application in future battery management systems (BMS), the model is then implemented on the AURIX TC4x from Infineon Technologies AG.